Detecting introgression despite phylogenetic ... - Wiley Online Library

DR. ELIZABETH BECKMAN (Orcid ID : 0000-0002-8303-2475) : From the Cover

Accepted Article

Article type

Detecting introgression despite phylogenetic uncertainty: the case of the South American siskins.

Running Head: Detecting introgression despite phylogenetic uncertainty

Elizabeth J. Beckman1,2,3, Phred M. Benham1, Zachary A. Cheviron1, Christopher C. Witt2 1

Division of Biological Sciences

University of Montana Missoula, Montana, USA 2

Department of Biology and Museum of Southwestern Biology

University of New Mexico Albuquerque, New Mexico, USA 3

Corresponding author: [email protected]

Key words: introgression, phylogenomics, molecular evolution, rapid radiation, Andes, Spinus

This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/mec.14795 This article is protected by copyright. All rights reserved.

Accepted Article

ABSTRACT: Genetic introgression among closely related species is a widespread phenomenon across the

Tree of Life and could be an important source of adaptive variation during early stages of diversification. In particular, genomic studies have revealed that many rapidly radiating clades tend to have complex, reticulate evolutionary histories. Although rapid radiations appear to be susceptible to introgression, they present special challenges for its detection because formal tests require accurate phylogenies, and paradoxically, introgression itself may obscure evolutionary relationships. To address this methodological challenge, we assessed introgression in a recent, rapid avian radiation in the Andes, the South American siskins (Spinus). Using ~45,000 SNPs, we estimated the Spinus phylogeny using multiple analytical approaches and recovered four strongly conflicting topologies. We performed a series of complimentary introgression tests that included valid tests for each of the likely species trees. From the consilience of test results, we inferred multiple introgression events among Andean Spinus in a way that was robust to phylogenetic uncertainty in the species tree. Positive tests for introgression were corroborated by independent population structure and ancestral assignment analyses, as well as a striking geographic pattern of mitochondrial haplotype-sharing among species. The methodological approach we describe could be applied using any genome-wide data, including SNP data, for clades without fully resolvable species trees. Our discovery of multiple introgression events within the Andean radiation of Spinus siskins is consistent with an emerging paradigm, that introgression tends to accompany the early stages of diversification.

This article is protected by copyright. All rights reserved.

Accepted Article

INTRODUCTION: Introgression, the transfer of genetic material from one independent lineage into the gene

pool of another (Mallet 2005), commonly occurs between closely related species pairs across the Tree of Life (Payseur & Rieseberg 2016). Introgression can impact functional traits (Lewontin & Birch 1966; Miao et al. 2016), and shape adaptive divergence by altering standing genetic variation (Parsons et al. 2011; Pease et al. 2016; Grant & Grant 2016; Winger 2017). Within rapid radiations where nascent species barriers may be challenged early in diversification, introgression may be of particular importance either as a creative or homogenizing force (Seehausen 2004). Genomic studies in several rapid radiations in plants, arthropods, and vertebrates have revealed reticulate evolutionary histories with introgression occurring among multiple species within a clade (Dasmahapatra et al. 2012; Cui et al. 2013; Keller et al. 2013; Gompert et al. 2014; Fontaine et al., 2015; Lamichhaney et al. 2015; Alexander et al. 2016; Meyer et al. 2016; Vargas et al. 2017). Within recent, speciose clades, alleles with major effects have introgressed with consequences on diet (Lamichhaney et al. 2015; Grant & Grant 2016; Richards & Martin 2017), behavioral reproductive isolation (Genner & Turner 2012; Keller et al. 2013; Meyer et al. 2016; Meier et al. 2017), predation risk (Dasmahapatra et al. 2012; Wallbank et al. 2016), and geographic range (Fontaine et al. 2015; Wen et al. 2016a). All told, these results suggest that introgression may be a typical and important characteristic of rapid radiations (Seehausen 2004; Mallet et al. 2015). However, nuclear DNA (nDNA) introgression does not always occur when closely related lineages come into contact (MeloFerreira et al. 2014; Good et al. 2015), and introgression has been formally assessed in only a small fraction of the rapid radiations found across the Tree of Life. Defining the frequency and predictability of introgression in diverse lineages will be essential for understanding the early stages of diversification.


Studying introgression within rapid radiations is challenging. Formal introgression tests

Accepted Article

require accurate phylogenies (Patterson et al. 2012); paradoxically, introgression itself may obscure the phylogenetic relationships within recent radiations. Rapid radiations have two potential sources of gene heterogeneity: incomplete lineage sorting (ILS) (Maddison 1997; Edwards 2009; Linkem et al. 2016) and introgression (Yu et al. 2014; Pease & Hahn 2015; Solís-Lemus et al. 2016). The presence of incomplete lineage sorting, or retention of ancestral variation across divergence events, is more probable when speciation events occur in close succession (Maddison 1997). Coalescent-based phylogenetic methods were developed to address this problem, but introgression violates the assumptions of widely-used traditional and coalescent-based phylogenetic approaches (Yu et al. 2012, 2014; Edwards et al. 2016; Solís-Lemus et al. 2016). When introgression is rampant, the true species tree may be represented in only a small fraction of loci across the genome (Fontaine et al. 2015); in other cases, a large proportion of the genome must be sequenced to estimate a statistically robust phylogeny (Cui et al. 2013; Lamichhaney et al. 2015). Phylogenetic uncertainty in rapid radiations may remain due to either source of gene-tree heterogeneity, even with genome-scale data (Suh et al. 2015; Linkem et al. 2016; Novikova et al. 2016). Often the extent and source of phylogenetic uncertainty is only revealed with exhaustive phylogenetic analyses (Smith et al. 2015; Crowl et al. 2017; Shen et al. 2017). Further, conventional methods for assessing phylogenetic support may over-represent the confidence for a specific reconstruction when large amounts of data are available (Jarvis et al. 2014; Smith et al. 2015; Suh 2016). To elucidate evolutionary history despite the presence of ILS and introgression, researchers have taken several different approaches, including: comparing the fit of genomic data when migration is allowed or excluded in population models that incorporate colonization history derived from non-genetic information (Alexander et al. 2016); modifying existing introgression tests for more complex scenarios (Eaton et al. 2015; Pease & Hahn 2015; Morales et al. 2017), and explicitly incorporating introgression into phylogenetic networks based on gene trees (Yu et al. 2012, 2014; Meyer et al. 2016; Solís-Lemus & Ané 2016; Wen et al. 2016a; Wen et al. 2016b). Despite this progress, a simple, general approach that can be


applied to any unresolved phylogeny, including clades without genomic resources, is needed to

Accepted Article

establish the extent and significance of introgression in the early stages of diversification. Here, we recommend a set of existing analyses that together allow conservative inference of

introgression events even when the phylogeny remains unresolved. This approach could be broadly applicable to a wide range of organisms due to its simple data requirements: modest intraspecific sampling for the focal taxa, a nuclear genome-wide SNP or sequence dataset, and a mitochondrial or chloroplast DNA sequence dataset. Appropriate SNP datasets can be produced by popular sequencing protocols (Leaché & Oaks 2017; Ree and Hipp 2015) and de novo assembly methods (Catchen et al. 2011; Eaton 2014) that are affordable and accessible even for non-model taxa without genomic resources. With these type of data, we propose the following steps: (1) Explore population structure by using nDNA sequence divergence and population allele frequencies. In clades impacted by incomplete lineage sorting and/or introgression, different analytical methods may yield discordant phylogenetic hypotheses. (2) Test for introgression by employing a suite of formal introgression tests based on site pattern and population allele frequencies; formal four-taxon introgression tests that are valid for each of the alternative phylogenetic hypotheses should be included. (3) Identify introgression events by looking for consensus across tests. (4) Compare nDNA introgression patterns with the evolutionary history of a mitochondrial or chloroplast genome. We applied this approach to test introgression in a rapid radiation of Neotropical finches,

the South American siskins (Spinus). The South American clade of Spinus, with 11 endemic species on the continent, is recent (~0.55 million years old), diversified rapidly (Beckman & Witt 2015), and has maximum diversity in the high Central Andes where three to four Spinus species co-occur over the course of the year (Ridgely & Tudor 1989; Fjeldså & Krabbe 1990). There is extensive mitochondrial haplotype sharing among sympatric Andean siskins; for example, all Andean individuals of the widespread Spinus magellanicus possessed mitochondrial DNA (mtDNA) haplotypes identical to either S. atratus or S. crassirostris, two high-elevation restricted taxa (Beckman & Witt 2015).


Reticulate evolutionary histories have been well-documented in a similarly rapid granivorous

Accepted Article

passerine bird radiation, the Darwin’s finches, (Grant et al. 2005; Lamichhaney et al. 2015; Grant & Grant 2016; Lamichhaney et al. 2016), and may be characteristic of rapid Andean plant radiations (Pease et al. 2016; Vargas et al. 2017). However, it is possible that mtDNA haplotype sharing in Andean Spinus could be the product of incomplete lineage sorting alone, and that species barriers have been maintained in sympatry. Testing for introgression in Spinus is an appropriate challenge with which to demonstrate the recommendations we propose, and our study provides potential insight into the biogeography and genetics of reticulate evolutionary histories.

METHODS: Taxonomic Sampling: We used museum collections and field expeditions to compile frozen tissue samples for 148

Spinus individuals including: three species restricted to high elevation in the Andes, S. atratus (N=8), S. crassirostris (N=16), and S. uropygialis (N=20); one widespread, polytypic species, S. magellanicus, sampled broadly throughout its western range [Ecuador (N=3), Peru (N=89), Bolivia (N=4) and northern Argentina (N=4)]. These taxa are referred to as Andean Spinus hereafter. We included S. cucullatus (N=4) sampled from Guyana as an outgroup. Previous work demonstrated this sampling should accurately represent mtDNA genetic diversity for these species (Beckman & Witt 2015).

DNA Sequencing: Genotyping-By-Sequencing: We extracted DNA from muscle tissue samples with the Qiagen DNeasy Blood and Tissue kit

(Qiagen, Valencia, CA). We used a modified, double-digest, genotyping-by-sequencing (GBS)


approach (Parchman et al. 2012) to sequence tens of thousands of loci from across the genome. We

Accepted Article

chose 40 individuals with high quality extracts for GBS (Table 1) to represent the geographic, taxonomic and mtDNA diversity present in Spinus (Beckman & Witt 2015). We included four individuals per species of S. atratus, S. crassirostris and S. uropygialis, and one S. cucullatus. We sequenced S. magellanicus from four regions: central Peru including the departments of Lima and Ancash, Peru (N=16), the department of Cusco, Peru (N=5), Cochabamba, Bolivia (N=3), and northern Argentina (N=3). We followed Parchman et al. (2012) to generate sequencing libraries using a size-selection

window of 500-600 bp. We pooled individual libraries in equimolar ratios to produce the pooled product which was sequenced on a single flow-cell lane of Illumina HiSeq 2500 at the W. M. Keck Sequencing Center at the University of Illinois, Urbana-Champaign. Sequencing resulted in >200 million single-end 100 bp reads with average Phred scores over 30. We used the Stacks pipeline v1.35 (Catchen et al. 2011; Catchen et al. 2013) to process the

GBS raw reads into de novo loci. Stacks recovers homologous loci between different samples without the aid of a genome, and it is effective across a range of divergence levels. Its flexibility is due to three user-designated parameters; the raw read depth required to assemble a locus (m), the maximum mismatches permitted within a locus in an individual (M) and the maximum mismatches allowed between individuals within a single locus (n). The appropriate values of these parameters vary based on the divergence expected between individuals, the depth of sequencing and the sequence error rate (Mastretta-Yanes et al. 2015). Based on previous work (Catchen et al. 2013; Harvey et al. 2015; Mastretta-Yanes et al. 2015) and preliminary analysis of m with a single individual, we set m=4 to maximize the number of high-confidence loci recovered. To test the robustness of our assembly to variation in assembly parameters, we processed the data through Stacks under five conditions: (B1) n=1, M=2, (B2) n=2, M=2, (B3) n=2, M=3, (B4) n=3, M=3, (B5) n=3, M=4. B1 had most restrictive assembly criteria (97% sequence similarity), B5 was most lenient (93%


sequence similarity). We eliminated potentially confounded loci from highly repetitive regions

Accepted Article

during locus construction (--max_locus_stacks = 3), and by filtering with rxstacks (--lnl_lim, -conf_filter) (Catchen et al. 2013). We evaluated assemblies B1 through B5 by calculating the average number of unique alleles

per locus for each. Under inappropriately strict mismatch parameters, a true homologous locus with distinct, divergent alleles may be artificially split, resulting in a lower value of average unique alleles per locus (Harvey et al. 2015). Comparable values of the average unique alleles per locus between assemblies indicate that the loci constructed in each are robust to parameter changes. To minimize the number of non-homologous loci that were incorrectly combined, we chose the most restrictive assembly, assembly B3, from those with similar unique alleles per locus to use for all subsequent analyses. To control the amount of missing data, we used the parameters r and p in the Stacks

populations program to filter the SNP dataset; r defines the proportion of individuals required per population for a locus to be considered, whereas p refers the number of populations required for a locus (Catchen et al. 2013). We treated each species as a taxonomic unit in the populations program except for S. magellanicus. Informed by previous work (Beckman & Witt 2015) and preliminary analyses, we analyzed S. magellanicus as three populations defined by latitude: 8º–14º S, hereafter 'Peru' (N = 21); 15º–21º S, hereafter 'Bolivia' (N = 3), and 22º–28º S, hereafter 'Argentina' (N = 3). These designations resulted in seven Spinus lineages. Mitochondrial Genes: For all 148 Spinus samples, excepting 53 S. magellanicus, we amplified and sequenced the

mitochondrial genes cytochrome b (cytb), NADH dehydrogenase II (ND2) and NADH dehydrogenase III (ND3) as in Beckman & Witt (2015). We found ND3 was sufficient to identify major mtDNA clades and sequenced the gene for the remaining 53 S. magellanicus. Sequencing was conducted at the


University of New Mexico Molecular Core Facility (Albuquerque, NM). We edited chromatograms by

Accepted Article

eye in Sequencher 4.10.1 (GeneCodes Corp., Ann Arbor, MI), aligned loci with Muscle v3.7 (Edgar 2004) on the CIPRES portal (Miller et al. 2010) and confirmed the alignments by eye in MacClade v4.08 (Maddison & Maddison 2005).

Population Structure: We assessed the population genetic structure of Andean Spinus with the GBS data by

performing a principal component analysis (PCA) with the R-package adegenet v2.0.1 (Jombart & Amhed, 2011) in R v3.2.4 (R Core Team, 2016). The input included all variant sites from the best assembly with r = 0.5. Since the initial PCA separated only two lineages from the all other taxa, we ran a second PCA omitting the two divergent taxa to assess structure within the remaining populations. We used the model-based likelihood clustering algorithm implemented in Admixture v1.3

(Alexander et al. 2009) and the GBS data to assign Andean Spinus individuals to populations and admixture fractions to individuals without a priori population assignment. The input included the first SNP from every variable locus in assembly B3 with r = 0.5 and p = 6. To identify the appropriate number of populations (k) for the data, we calculated the cross-validation error (CV error) for k equal to 1-10 and identified k with the lowest values (Alexander et al. 2009).

Phylogenetics: We estimated a maximum likelihood (ML) phylogenetic tree from the concatenated GBS

data using the hybrid, MPI/PThreads version of RAxML v8.1.17 (Pfeiffer & Stamatakis 2010; Stamatakis 2014). The computation efficiency of RAxML permitted the inclusion of tens of thousands


of loci for all 40 individuals sequenced; the inclusion of more sites can reveal strong support for

Accepted Article

nodes that are weakly supported by data subsets and other phylogenetic methods (Jarvis et al. 2014). The rapid computational time of RAxML also allowed a thorough exploration of the effect of missing data on the phylogeny, a critical component to the analysis of SNP datasets with large amounts of missing data (Davey et al. 2011). To assess the consequences of missing data on topology and node support, we constructed a series of SNP datasets using assembly B3 in which we varied the missing data parameters r from 0.5 to 0.75, and p from 4 to 7. We concatenated all variant sites within a dataset, and used the Felsenstein invariant site correction with molecular evolution model ASC_GTRGAMMA in RAxML (Leaché et al. 2015) with 500 bootstrap replicates. The number of invariant sites present in phylogenetic construction will impact the branch lengths and therefore the likelihood of each tree (Lewis 2001; Leaché et al. 2015). The Felsenstein invariant site correction allays this problem and reduces computational time compared to an analysis with complete, concatenated loci. For the final phylogeny, hereafter RAxML-B3 tree, we selected the r and p combination which minimized missing data and maximized phylogenetic support for internal nodes. We estimated the RAxML-B3 phylogeny using concatenated, complete loci (89 bp/locus), GTRGAMMA and 500 bootstrap replicates. We generated a phylogenetic tree using the multi-species coalescent with the program

SNAPP (Bryant et al. 2012) and the GBS data. In this approach, unlinked, bi-allelic sites have independent gene histories that are constrained by the ultimate species tree (Edwards 2009; Bryant et al. 2012). Gene heterogeneity is attributed solely to differences in the time of coalescence and incomplete lineage sorting (Bryant et al. 2012). For this analysis, we maximized the number of sites covered across individuals by selecting 20 high coverage individuals including S. cucullatus and 2-6 individuals of the remaining populations (Set E, Table 1). We extracted the first variant site from each locus that was present in >66% of individuals in each population (r = 0.66, p = 7). To reduce computational time, we binned SNPs into 3 equal sized datasets and ran each in a separate SNAPP run. We grouped S. magellanicus from Bolivia and Argentina, hereafter S. magellanicus Bol. Arg.,


into a single operational taxonomic unit analysis based on the RAxML analyses. We estimated the

Accepted Article

mutation rates U and V within BEAUti 2 (Bouckaert et al. 2014) from the data and used the default prior specifications. Each dataset was run for 3 million generations with a 10% burnin. We used Tracer v1.6 (Rambaut et al. 2014) to evaluate convergence. Once we established concordance among independent runs, we combined the trees and evaluated support for individual nodes in Densitree v2.2.4 (Bouckaert & Heled 2014). Lastly, we used a quartet inference approach with SVDquartets (Chifman & Kubatko 2014,

2015), implemented in PAUP* v4.0a147 (Swofford 2002). This coalescent method calculates the best tree topology by identifying a valid split across all loci for each quartet. The quartet trees are assembled with a variant of Quartet FM (Reaz et al. 2014) to create the final phylogeny. Again, we maximized the number of sites shared among individuals in SVDquartets by including 29 individuals (Set C, Table 1) with filtering parameters r = 0.66, p = 6. Note, we used p = 6 since unlike SNAPP, SVDquartets will permit a site to be completely missing on one population. Changing the population parameter p from 7 to 6 allowed us to include more individuals and sites in the SVDquartet analysis than the SNAPP analysis. We searched all possible quartets for all analyses. We estimated "lineage trees", where each individual is treated as a terminal branch, and "species trees", where members of each population are defined a priori. For individuals sequenced for mitochondrial genes cytb, ND3, and ND2, we ran

PartitionFinder 2 (Lanfear et al. 2016; Stamatakis 2014) using the greedy search algorithm (Lanfear et al. 2012), linked branch lengths and the Akaike’s information criterion (AIC) to select an appropriate partition and evolutionary model scheme for phylogenetic analysis. Using the evolutionary model GTRGAMMA, and four partitions (cytb codon 1; ND3 and ND2 codon 1; codon 2; and codon 3), we estimated a maximum likelihood phylogeny from the mitochondrial data using RAxML v7.2.7 (Stamatakis 2014) on the Cipres Science Gateway v. 3.1 (Miller et al. 2010) with 500 bootstrap replicates.


Accepted Article

nDNA Introgression Tests: To test for introgression among Andean Spinus with the GBS data, we used three tree-based

methods: the ABBA/BABA test on fixed sites (Green et al. 2010), the ABBA/BABA test on polymorphic sites (Durand et al. 2011; Patterson et al. 2012), and the four-population test (Reich et al. 2009). These tests varied in their assumptions and power (Alexander et al. 2009; Patterson et al. 2012; Rheindt et al. 2014; Martin et al. 2015a). We defined 31 well-supported four-taxon trees (referred to as Trees 1-31 hereafter) from our RAxML-B3 phylogeny (Table 2); a subset of these trees was also valid for topologies produced by SNAPP and SVDquartets (Table 2). To understand how the signal of introgression might vary across the landscape (Eaton et al. 2015), we used the PCA results to delineate four geographic populations with distinct genetic variation within S. magellanicus: Central Peru (CP, from Lima and Ancash), Southern Peru (SP, from Cusco), Bolivia, and Argentina. We defined positive evidence for introgression as (1) a raw z-score >|2.0|, the equivalent to an uncorrected p-value of 100) for demographic parameters (e.g. theta, upon which the coalescent model depends). Consistent with this interpretation, ESS values did not increase with increasing sampling generations. The topology recovered by SVDquartets (Fig. S4) was similar to the SNAPP-S2 tree. The

support for the nodes in conflict with the RAxML-B3 and SNAPP-S1 topology was low in "lineage" trees (bs: 63 for S. crassirostris/S. uropygialis clade, 77 for S. magellanicus Bol. Arg. sister to S. magellanicus Peru/S. atratus) and high in "species" trees (bs: 91, 98). In this analysis, S. cucullatus was placed sister to the ancestor of S. crassirostris and S. uropygialis (Fig. 4). nDNA Introgression Tests: The formal introgression tests were based on 31 four-taxon trees, all of which were

compatible with our RAxML-B3 tree. Sixteen of the 31 trees were also consistent with the SNAPP-S1 tree; six were valid for the SNAPP-S2 tree and the SVDquartets tree (Table 2). With the 31 trees, we were able to conduct multiple tests for introgression between specific pairs of taxa; the number of SNPs differed among tests based on coverage per individual and variation within loci (Table S1). We found strong evidence for introgression between S. crassirostris and Peruvian S.

magellanicus (Table 2); this was true when S. magellanicus Peru (Trees 6-8) or S. magellanicus Central Peru (Trees 9-11) were used as the terminal taxon. Z-scores from the ABBA/BABA fixed tests for S. crassirostris and S. magellanicus from Peru ranged from 2.24 to 7.15; Trees 6, 7 10 and 11 passed the FDR threshold of 0.05. The evidence for introgression was robust for three tree


topologies and marginal for the fourth. The number of sites impacted the power of individual tests,

Accepted Article

and may explain the low z-scores for some tests. We also found strong evidence for introgression between S. magellanicus Peru and S.

uropygialis; relevant Trees 16-30 were valid for the RAxML and SNAPP-S1 phylogenies (Table 2). The signal of introgression was almost uniformly positive (z-score > 2.0) in introgression tests based on polymorphic sites (though not all passed the FDR threshold), but mostly absent in the ABBA/BABA test based on fixed sites. Z-scores were high in tests for introgression between S. uropygialis and S. magellanicus Southern Peru, and lower with all Peruvian S. magellanicus or S. magellanicus Central Peru, including some z-scores that did not pass the FDR threshold. This suggests that introgression patterns between S. uropygialis and Peruvian S. magellanicus was spatially structured, with higher incidence in Southern Peru. Likewise, there was modest evidence, valid for all tree topologies, that S. magellanicus from

Southern Peru has introgressed with S. atratus. Introgression was well-supported in 3 of 4 polymorphic site tests (Tree 31, Table 2), but not the fixed-difference ABBA/BABA test; this difference is likely a consequence of the few fixed sites available for the latter. Evidence for introgression between S. magellanicus Bolivia and S. atratus was marginal. Tree

4, valid for the RAxML tree only, had a z-score of 2.02, but did not pass the FDR threshold. There was no support in Test 3, which was valid for RAxML and SNAPP-S1 trees. There was no definitive evidence for introgression among the following: S. crassirostris and

S. magellanicus Bolivia or Bol. Arg. (Trees 12-15); S. atratus and S. uropygialis (Trees 12 and 13); or S. crassirostris and S. uropygialis (Trees 14 and 15). Results for the former two taxon-pairs were valid for the SNAPP-S1 tree; however, no tests apply to SNAPP-S2 or SVDquartets phylogenies. Further, the test results from Trees 16-30 firmly reject introgression between S. magellanicus Bol. Arg. and S. magellanicus Peru.


Accepted Article

Mitochondrial DNA: We sequenced the genes ND3, ND2, and cytb for 95 Spinus samples and trimmed each to

416 bp, 777 bp and 753 bp, respectively. The concatenated mtDNA phylogeny revealed five wellsupported monophyletic clades corresponding to Spinus crassirostris (bs 84), S. uropygialis (bs 60), S. atratus (bs 97), S. cucullatus (bs 100), and S. magellanicus from Argentina (bs 100, Fig. 5A); individuals from each of these populations were found only in their respective clade. However, S. magellanicus from Peru and Bolivia were found in the clades corresponding to S. crassirostris and S. atratus (Fig. 5A). The above results were recapitulated with ND3 alone; thus, we used ND3 to characterize 100 samples of S. magellanicus into S. crassirostris, S. atratus or S. magellanicus Argentina clades. In S. magellanicus individuals, the S. crassirostris haplotype group predominated on the western slope of the Andes in Ecuador and Peru at all elevations (Fig. 5B). The S. atratus haplotype group increased in frequency at high elevations near Cusco, Peru, and was fixed in Cochabamba, Bolivia at 2470 m. In the foothills and plains southeast of the central Andes, we found only S. magellanicus Argentina haplotypes (Fig. 5B). DISCUSSION: We documented introgression in a rapid radiation, the South American siskins, despite

persistent uncertainty within the true species tree. Although we applied a suite of phylogenetic methods, some key internal nodes of the Spinus phylogeny remained unresolved. While considering a set of alternative phylogenetic trees, we tested for nDNA introgression among Spinus taxa using phylogeny-based tests for both fixed and polymorphic loci. We corroborated these results with independent population structure analyses for the nuclear and mitochondrial genomes. The preponderance of the results was consistent with multiple introgression events occurring among Spinus species, which likely contributed to the conflicting phylogenetic signals that we detected.


We report definitive evidence for interspecific introgression between each of the following

Accepted Article

pairs of taxa: (1) S. magellanicus Peru and S. crassirostris; (2) S. magellanicus Peru and S. uropygialis; (3) S. magellanicus Southern Peru and S. atratus; (4) S. magellanicus Bolivia and S. atratus. Introgression between each pair of taxa was supported by concordant results across all or a subset of the following: tree-based introgression tests, ancestral assignment analyes, population structure analyses, and mtDNA haplotype distribution patterns (Table 2; Fig. 6). For taxon pairs 1 and 3, support from tree-based inrogression tests were valid for all major tree topologies (Fig. 6a-d); evidence supporting introgression in taxon pair 2 was valid for the RAxML-B3 and the SNAPP-S1 trees (Fig. 6a-b); and evidence for introgression in taxon pair 4 applied to the RAxML-B3 tree only (Fig. 6a). Additional sampling of Spinus species or populations would be important to design appropriate, independent introgression tests for pairs 2 and 4 given the true species tree resembled either the SNAPP-S2 or SVDQuartet topologies. One caveat to our findings is that we were unable to formally assess multiple introgression

events in a single analysis. Thus, it is possible that the test statistic for a specific four-taxon tree may have been confounded by introgression between one focal species and a divergent lineage not included in the test, a so-called ghost taxon (Durand et al 2011). In this case, we would expect a noisy signal of introgression and more false positives overall (Pease & Hahn 2015). We think it is unlikely that our formal introgression tests were impacted by this problem since (1) two of our introgression taxon pairs are supported by multiple formal tests consisting of a unique set of four taxa, (2) we used a multiple test correction to minimize the overall false positive rate, and (3) each introgression taxon pair we recovered involved sympatric populations. The phylogenetic uncertainty we report within Spinus is a familiar problem encountered for

recent, rapid radiations (Dasmahapatra et al. 2012; Cui et al. 2013; Campagna et al. 2015; Fontaine et al. 2015; Linkem et al. 2016; Pease et al. 2016; McVay et al. 2017). Introgression likely contributed to phylogenetic uncertainty in Spinus, and the difficulty of resolving a bifurcating tree may have been


further compounded by conditions favorable for incomplete lineage sorting (Kubatko et al. 2007;

Accepted Article

Linkem et al. 2016). Recent studies on recalcitrant phylogenomic problems suggest that wellsupported strictly bifurcating trees should be considered with caution when there is evidence of conflict within the datasets (Suh 2016; Shen et al. 2017; Crowl et al. 2017; Thom et al. 2018). Instead of accepting high nodal support as conclusive, some authors have argued for routine investigation into the sources of gene conflict (Shen et al. 2017), and novel approaches for assessing node support (Smith et al., 2015; Arcila et al., 2017). For recently diverged taxa, a successful approach to understand evolutionary history may be

to model alternative demographic scenarios (Alexander et al. 2016; Thom et al. 2018). However, population modeling is most effective when a small number of alternative hypotheses can be compared with high confidence that one scenario represents the true history. Even in relatively simple scenarios with only four taxa, it may be difficult to elect the best model (Thom et al. 2018). Model sophistication and computational costs increase with the number of species and alternative scenarios. Phylogenetic networks, which depict the linear evolution of populations through time as

well as the strength and direction of reticulation events between taxa, are a particularly promising solution to the phylogenetic uncertainty due to introgression (Than et al. 2008; Gerard et al. 2011; Yu et al. 2011; Liu et al. 2014); however, our results in Andean Spinus highlight two significant challenges in the future development of phylogenetic network analyses. First, to confidently resolve a subset of the evolutionary relationships within this shallow clade, we required >45,000 independent loci distributed across the genome; most 89-bp loci in our dataset contained 0-2 variable sites. Phylogenetic topology and nodal support was sensitive to missing data as well as the total number of sites. These results illustrate the challenge of finding phylogenetically informative sites between recently diverged taxa; this problem may be exacerbated in phylogenetic network approaches where the initial inputs are informative, independent gene trees or their derivatives


(Solís-Lemus & Ané 2016, Wen et al. 2016a; Wen et al. 2016b). Improvements in computational

Accepted Article

efficiency allow the network software SNaQ (Solis-Lemus & Ané 2016) to incorporate as many as 1000 gene trees; however, this is well below the number of loci we used here, and individual gene trees, even of longer loci, will tend to be poorly resolved in shallow divergences due to the paucity of variable sites per locus. Systematic study is needed to determine whether network analyses such as SNAQ can potentially utilize large numbers of poorly resolved gene trees. Second, we found that multiple species pairs within Andean Spinus have introgressed; this problem of multiple reticulation events has been reported in other rapid radiations as well (Lamichhaney et al. 2015; Fontaine et al. 2015). Developing network analyses that retain the power to infer reticulation in complex scenarios involving many species is a significant computational challenge (Solís-Lemus & Ané 2016). The ability to estimate the direction and magnitude of historic gene flow in clades that include many taxa, including parapatric and sympatric species pairs, would contribute to our understanding of how and when introgression occurs (Schumer et al. 2017a, Schumer et al. 2017b), as well as how it may contribute to adaptive radiations (Kagawa & Takimoto 2017; Lamichhaney et al. 2017). We found that sampling S. magellanicus across a broad geographic area was important for

providing insight about how introgression has occurred among Andean Spinus populations. For example, we found that S. magellanicus from Bolivia introgressed with S. atratus while genetically similar S. magellanicus from Argentina did not. Admixture results (Fig. 2, k=4) revealed two S. atratus individuals with a high proportion of S. magellanicus Bolivia-Argentina ancestry, suggesting that these individuals may have resulted from recent hybridization or backcrossing. Thus, introgression with S. atratus may be recent and localized to Bolivia rather than an ancient reticulation event. In contrast, the introgression between S. crassirostris and S. magellanicus from Peru does not appear to be spatially structured; this could indicate that introgression has occurred across the entire extent of Peru where these lineages are sympatric or that introgression between these two lineages was ancestral. Future work comparing the genomic patterns in these two cases could reveal whether introgression in Spinus reflects recent interactions or ancient ones.


Analyses of mtDNA variation revealed geographic structure associated with elevation; 9 out

Accepted Article

of 10 S. magellanicus that possessed S. atratus-like mtDNA were captured at elevations between 2500 m and 3800 m. S. atratus itself is restricted to elevations >3500 m while S. magellenicus is found from 0 to 4500 m (Ridgely & Tudor 1989). The clade consisting of S. magellanicus from Bolivia and Argentina (including birds collected as low as 340 m) was well supported by nDNA; yet, Bolivian S. magellanicus from 2550 m were fixed for the S. atratus mtDNA haplotype. mtDNA is a potential target for natural selection with respect to elevation due to its role in cellular respiration (Mishmar et al. 2003; Ballard & Whitlock 2004; Blier et al. 2006; Boratyński et al. 2011), thermal tolerance (Gering et al. 2009) and the reduction of reactive oxygen species under hypoxic conditions (Scott et al. 2011). Local adaptation of mtDNA to elevation is suspected to have caused fine-scale mtDNA structure along the west slope of the Central Andes in the widespread sparrow Zonotrichia capensis (Cheviron & Brumfield 2009). The distribution of the S. atratus mtDNA haplotype in S. magellanicus is consistent with elevationally driven adaptive introgression. In light of the circumstantial evidence cited above and the theoretical results of Bonnet et al. (2017), we tentatively conclude adaptive introgression is likely explanation for mitonuclear discordance in Spinus.

Conclusions: Significant challenges remain for rigorously identifying introgression in natural populations.

Here, we employed a practical approach for evaluating phylogenetic conflict and overcoming it to achieve robust inference of introgression among closely related species; our recommendations are broadly applicable to non-model organisms due to the simple data requirements and the capacity to deal with phylogenetic uncertainty. Our discovery of multiple introgression events within the Andean radiation of Spinus siskins is consistent with an emerging paradigm, that introgression tends to accompany the early stages of diversification (Dasmahapatra et al. 2012; Fontaine et al. 2015; Martin et al. 2015b; Grant & Grant 2016; Pease et al. 2016). Conservative inference of genetic


introgression is essential because of numerous recent reports that introgressed genes can have

Accepted Article

profound consequences for trait variation (Lamichhaney et al. 2016; Richards & Martin 2017), and diversification (Barrera-Guzmán et al. 2017; Kagawa & Takimoto 2017, Lamichhaney et al. 2017).

ACKNOWLEDGEMENTS: For tissue samples, we thank: the University of Washington Burke Museum (Sharon Birks),

the Smithsonian Institute (Christopher Milensky, Michael Braun); the Academy of Natural Sciences, Philadelphia (Nate Rice); the Field Museum of Natural History (David Willard, John Bates); the Cornell University Museum of Vertebrates (Kimberly Bostwick) and the Museum of Southwestern Biology, University of New Mexico as well as the many prescient collectors. We gratefully acknowledge funding from NSF (DEB 1146491), the Frank M. Chapman Memorial Fund at the American Museum of Natural History, and the American Ornithological Society. Last, we thank Bryan Carstens, Michael Gruenstaeudl, and two anonymous reviewers for their valuable comments.


REFERENCES:

Accepted Article

Alexander, A. M., Su, Y.-C., Oliveros, C. H., Olson, K. V., Travers, S. L., & Brown, R. M. (2016). Genomic data reveals potential for hybridization, introgression, and incomplete lineage sorting to confound phylogenetic relationships in an adaptive radiation of narrow-mouth frogs. Evolution, 85, 10–15. https://doi.org/10.1111/evo.13133 Alexander, D. H., Novembre, J., & Lange, K. (2009). Fast Model-Based Estimation of Ancestry in Unrelated Individuals. Genome Research, 19, 1655–1664. https://doi.org/10.1101/gr.094052.109

Arcila, D., Ortí, G., Vari, R., Armbruster, J., Stiassny, M., Kyung, K., Sabaj, M., Lundberg, J., Revell, L. & Betancur-R, R. (2017). Genome-wide interrogation advances resolution of recalcitrant groups in the tree of life. Nature Ecology & Evolution 1: 0020. Doi:10.1038/s41559-016-0020 Ballard, J. W. O., & Whitlock, M. C. (2004). The incomplete natural history of mitochondria. Molecular Ecology, 13, 729–744. https://doi.org/10.1046/j.1365-294X.2003.02063.x

Barrera-Guzmán, A. O., Aleixo, A., Shawkey, M. D., & Weir, J. T. (2018). Hybrid speciation leads to novel male secondary sexual ornamentation of an Amazonian bird. Proceedings of the National Academy of Sciences. 115: E218E225. https://doi.org/10.1073/pnas.1717319115 Beckman, E. J., & Witt, C. C. (2015). Phylogeny and biogeography of the New World siskins and goldfinches: Rapid, recent diversification in the Central Andes. Molecular Phylogenetics and Evolution, 87, 28–45. https://doi.org/10.1016/j.ympev.2015.03.005 Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society B. 57, 289-300. Retrieved from http://www.jstor.org/stable/2346101 Benjamini, Y., & Yekutieli, D. (2001). The Control of the False Discovery Rate in Multiple Testing under Dependency. The Annals of Statistics, 29, 1165–1188. Retrieved from http://www.jstor.org/stable/2674075 Blackmon, H. (2016). R package: evobiR. Blier, P. U., Breton, S., Desrosiers, V., & Lemieux, H. (2006). Functional Conservatism in mitochondrial evolution: Insight from hybridization of Arctic and Brook Charrs. Journal of Experimental Zoology. Part B, Molecular and Developmental Evolution, 306, 425–432. Bonnet, T., Leblois, R., Rousset, F., & Crochet, P. A. (2017). A reassessment of explanations for discordant introgressions of mitochondrial and nuclear genomes. Evolution, 1–45. https://doi.org/10.1111/evo.13296 Boratyński, Z., Alves, P., Berto, S., Koskela, E., Mappes, T., & Melo-Ferreira, J. (2011). Introgression of mitochondrial DNA among Myodes voles: consequences for energetics? BMC Evolutionary Biology, 11, 355. https://doi.org/10.1186/1471-2148-11-355


Bouckaert, R., & Heled, J. (2014). DensiTree 2: Seeing Trees Through the Forest. bioRxiv, 12401. https://doi.org/10.1101/012401

Accepted Article

Bouckaert, R., Heled, J., Kühnert, D., Vaughan, T., Wu, C., Xie, D., Suchard, M. A., Rambaut, A., & Drummond, A. J. (2014). BEAST 2: A software platform for Bayesian evolutionary analysis. PLoS Comput Biol 10: e1003537. https://doi.org/10.1371/journal.pcbi.1003537 Bryant, D., Bouckaert, R., Felsenstein, J., Rosenberg, N. A., & Roychoudhury, A. (2012). Inferring species trees directly from biallelic genetic markers: Bypassing gene trees in a full coalescent analysis. Molecular Biology and Evolution, 29, 1917–1932. https://doi.org/10.1093/molbev/mss086 Campagna, L., Gronau, I., Silveira, L. F., Siepel, A., & Lovette, I. J. (2015). Distinguishing noise from signal in patterns of genomic divergence in a highly polymorphic avian radiation. Molecular Ecology, 24, 4238–51. https://doi.org/10.1111/mec.13314 Catchen, J., Hohenlohe, P. A., Bassham, S., Amores, A., & Cresko, W. A. (2013). Stacks: An analysis tool set for population genomics. Molecular Ecology, 22, 3124–3140. https://doi.org/10.1111/mec.12354 Catchen, J. M., Amores, A., Hohenlohe, P., Cresko, W., Postlethwait, J. H., & De Koning, D.-J. (2011). Stacks: Building and Genotyping Loci De Novo From Short-Read Sequences. Genes|Genomes|Genetics, 1, 171–182. https://doi.org/10.1534/g3.111.000240 Cheviron, Z. A., & Brumfield, R. T. (2009). Migration-selection balance and local adaptation of mitochondrial haplotypes in Rufous-Collared Sparrows (Zonotrichia Capensis) along an elevational gradient. Evolution, 63, 1593–1605. https://doi.org/10.1111/j.15585646.2009.00644.x Chifman, J., & Kubatko, L. (2014). Quartet Inference from SNP Data Under the Coalescent Model. Bioinformatics, 30, 3317–3324. https://doi.org/10.1093/bioinformatics/btu530 Chifman, J., & Kubatko, L. (2015). Identifiability of the unrooted species tree topology under the coalescent model with time-reversible substitution processes, site-specific rate variation, and invariable sites. Journal of Theoretical Biology, 374, 35–47. https://doi.org/10.1016/j.jtbi.2015.03.006 Crowl, A. A., Myers, C., & Cellinese, N. (2017). Embracing discordance: Phylogenomic analyses provide evidence for allopolyploidy leading to cryptic diversity in a Mediterranean Campanula (Campanulaceae) clade. Evolution, 71, 913–922. https://doi.org/10.1111/evo.13203 Cui, R., Schumer, M., Kruesi, K., Walter, R., Andolfatto, P., & Rosenthal, G. G. (2013). Phylogenomics Reveals Extensive Reticulate Evolution In Xiphophorus Fishes. Evolution, 67, 2166–2179. https://doi.org/10.1111/evo.12099 Dasmahapatra, K. K., Walters, J. R., Briscoe, A. D., Davey, J. W., Whibley, A., Nadeau, N. J., Zimin, A. V., Hughes, D. S. T., Ferguson, L. C., Martin, S. H., Salazar, C., Lewis, J. J., Adler, S., Ahn, S., Baker, D. A., Baxter, S. W., Chamberlain, N. L., Chauhan, R., Counterman, B. A., Dalmay, T., Gilbert, L. E., Gordon, K., Heckel, D. G., Hines, H. M., Hoff, K. J., Holland, P. W. H., Jacquin-Joly, E., Jiggins, F.


Accepted Article

M., Jones, R. T., Kapan, D. D., Kersey, P., Lamas, G., Lawson, D., Mapleson, D., Maroja, L. S., Martin, A., Moxon, S., Palmer, W. J., Papa, R., Papanicolaou, A., Pauchet, Y., Ray, D. A., Rosser, N., Salzberg, S. L., Supple, M. A., Surridge, A., Tenger-Trolander, A., Vogel, H., Wilkinson, P. A., Wilson, D., Yorke, J. A., Yuan, F., Balmuth, A. L., Eland, C., Gharbi, K., Thomson, M., Gibbs, R. A., Han, Y., Jayaseelan, J. C., Kovar, C., Mathew, T., Muzny, D. M., Ongeri, F., Pu, L., Qu, J., Thornton, R. L., Worley, K. C., Wu, Y., Linares, M., Blaxter, M. L., Ffrench-Constant, R. H., Joron, M., Kronforst, M. R., Mullen, S. P., Reed, R. D., Scherer, S. E., Richards, S., Mallet, J., McMillan, W. O. & Jiggins, C. D. (2012). Butterfly genome reveals promiscuous exchange of mimicry adaptations among species. Nature, 487, 94–98. https://doi.org/10.1038/nature11041 Davey, J. W., Hohenlohe, P. A., Etter, P. D., Boone, J. Q., Catchen, J. M., & Blaxter, M. L. (2011). Genome-wide genetic marker discovery and genotyping using next-generation sequencing. Nature Reviews Genetics, 12: 499-510. Doi:10.1038/nrg3012 Dowling, T. E., & DeMarais, B. D. (1993). Evolutionary significance of introgressive hybridization in cyprinid fishes. Nature, 362, 444–446. https://doi.org/10.1038/362444a0 Durand, E. Y., Patterson, N., Reich, D., & Slatkin, M. (2011). Testing for ancient admixture between closely related populations. Molecular Biology and Evolution, 28, 2239–2252. https://doi.org/10.1093/molbev/msr048 Eaton, D. 2014. PyRAD: assembly of de novo RADseq loci for phylogenetic analyses. Bioinformatics, 30: 1844–1849, https://doi.org/10.1093/bioinformatics/btu121 Eaton, D. A. R., Hipp, A. L., González-Rodríguez, A., & Cavender-Bares, J. (2015). Historical introgression among the American live oaks and the comparative nature of tests for introgression. Evolution, 69, 2587–2601. https://doi.org/10.1111/evo.12758 Edwards, S. V. (2009). Is a new and general theory of molecular systematics emerging? Evolution, 63, 1–19. https://doi.org/10.1111/j.1558-5646.2008.00549.x Edwards, S. V., Xi, Z., Janke, A., Faircloth, B. C., McCormack, J. E., Glenn, T. C., Zhong, B., Wu, S., Lemmon, E. M., Lemmon, A. R., Leaché, A. D., Liu, L., & Davis, C. C. (2016). Implementing and testing the multispecies coalescent model: A valuable paradigm for phylogenomics. Molecular Phylogenetics and Evolution, 94, 447–462. https://doi.org/10.1016/j.ympev.2015.10.027 Elshire, R. J., Glaubitz, J. C., Sun, Q., Poland, J. A., Kawamoto, K., Buckler, E. S., & Mitchell, S. E. (2011). A robust, simple genotyping-by-sequencing (GBS) approach for high diversity species. PLoS ONE, 6, e19379. https://doi.org/10.1371/journal.pone.0019379 Fjeldså, J., & Krabbe, N. K. (1990). Birds of the High Andes. Copenhagen: Apollo Books. Fontaine, M. C., Pease, J. B., Steele, A., Waterhouse, R. M., Neafsey, D. E., Sharakhov, I. V, Jiang, X., Hall, A.B., Catteruccia, F., Kakani, E., Mitchell, S. N., Wu, Y. C., Smith, H. A., Love, R. R., Lawniczak, M. K., Slotman, M. A., Emrich, S. J., Hahn, M. W., Besansky, N. J. (2015). Extensive introgression in a malaria vector species complex revealed by phylogenomics. Science, 347, 1258524– 1258524. https://doi.org/10.1126/science.1258524 Ford, A. G. P., Dasmahapatra, K. K., Rüber, L., Gharbi, K., Cezard, T., & Day, J. J. (2015). High levels of interspecific gene flow in an endemic cichlid fish adaptive radiation from an extreme lake environment. Molecular Ecology, 24, 3421–3440. https://doi.org/10.1111/mec.13247


Accepted Article

Gatesy, J., & Springer, M. S. (2014). Phylogenetic analysis at deep timescales: Unreliable gene trees, bypassed hidden support, and the coalescence/concatalescence conundrum. Molecular Phylogenetics and Evolution, 80, 231–266. https://doi.org/10.1016/j.ympev.2014.08.013 Genner, M. J., & Turner, G. F. (2012). Ancient Hybridization and Phenotypic Novelty within Lake Malawi’s Cichlid Fish Radiation. Molecular Biology and Evolution, 29, 195–206. https://doi.org/10.1093/molbev/msr183 Gerard, D., Gibbs, H. L., & Kubatko, L. (2011). Estimating hybridization in the presence of coalescence using phylogenetic intraspecific sampling. BMC Evolutionary Biology, 11, 291. https://doi.org/10.1186/1471-2148-11-291 Gering, E. J., Opazo, J. C., & Storz, J. F. (2009). Molecular evolution of cytochrome b in high- and lowaltitude deer mice (genus Peromyscus). Heredity, 102, 226–235. https://doi.org/10.1038/hdy.2008.124 Gompert, Z., Lucas, L. K., Buerkle, C. A., Forister, M. L., Fordyce, J. A., & Nice, C. C. (2014). Admixture and the organization of genetic diversity in a butterfly species complex revealed through common and rare genetic variants. Molecular Ecology, 23, 4555–4573. https://doi.org/10.1111/mec.12811 Good, J. M., Vanderpool, D., Keeble, S., & Bi, K. (2015). Negligible nuclear introgression despite complete mitochondrial capture between two species of chipmunks. Evolution, 69, 1961–1972. https://doi.org/10.1111/evo.12712 Grant, P. R., & Grant, B. R. (2016). Introgressive hybridization and natural selection in Darwin’s finches. Biological Journal of the Linnean Society, 117, 812–822. https://doi.org/10.1111/bij.12702 Green, R. E., Krause, J., Briggs, A. W., Maricic, T., Stenzel, U., Kircher, M., Patterson, N., Li, H., Zhai, W., Fritz, Markus H., Hansen, N. F., Durand, E. Y., Malaspinas, A., Jensen, J. D., Marques-Bonet, T., Alkan, C., Prüfer, K., Meyer, M., Burbano, H. A., Good, J. M., Schultz, R., Aximu-Petri, A., Butthof, A., Höber, B., Höffner, B., Siegemund, M., Weihmann, A., Nusbaum, C., Lander, E. S., Russ, C., Novod, N., Affourtit, J., Egholm, M., Verna, C., Rudan, P., Brajkovic, D., Kucan, Ž., Gušic, I., Doronichev, V. B., Golovanova, L. V., Lalueza-Fox, C., de la Rasilla, M., Fortea, J., Rosas, A., Schmitz, R.W., Johnson, P. L. F., Eichler, E. E., Falush, D., Birney, E., Mullikin, J. C., Slatkin, M., Nielsen, R., Kelso, J., Lachmann, M., Reich, D. & Pääbo, S. (2010). A Draft Sequence of the Neandertal Genome. Science, 328, 710–722. https://doi.org/10.1126/science.1188021 Harvey, M. G., Judy, C. D., Seeholzer, G. F., Maley, J. M., Graves, G. R., & Brumfield, R. T. Similarity thresholds used in short read assembly reduce the comparability of population histories across species. PeerJ. 3, e864v1. https://doi.org/10.7287/peerj.preprints.864v1 Jarvis, E. D., Mirarab, S., Aberer, A. J., Li, B., Houde, P., Li, C., Ho, Simon Y. W., Faircloth, B. C., Nabholz, B., Howard, J. T., Suh, A., Weber, C. C., da Fonseca, R. R., Li, J., Zhang, F., Li, H., Zhou, L., Narula, N., Liu, L., Ganapathy, G., Boussau, B., Bayzid, M. S., Zavidovych, V., Subramanian, S., Gabaldon, T., Capella-Gutierrez, S., Huerta-Cepas, J., Rekepalli, B., Munch, K., Schierup, M., Lindow, B., Warren, W. C., Ray, D., Green, R. E., Bruford, M. W., Zhan, X., Dixon, A., Li, S., Li, N., Huang, Y., Derryberry, E. P., Bertelsen, M. F., Sheldon, F. H., Brumfield, R. T., Mello, C. V., Lovell,


Accepted Article

P. V., Wirthlin, M., Schneider, M. P. C., Prosdocimi, F., Samaniego, J. A., Velazquez, A. M. V., Alfaro-Nunez, A., Campos, P. F., Petersen, B., Sicheritz-Ponten, T., Pas, A., Bailey, T., Scofield, P., Bunce, M., Lambert, D. M., Zhou, Q., Perelman, P., Driskell, A. C., Shapiro, B., Xiong, Z., Zeng, Y., Liu, S., Li, Z., Liu, B., Wu, K., Xiao, J., Yinqi, X., Zheng, Q., Zhang, Y., Yang, H., Wang, J., Smeds, L., Rheindt, F. E., Braun, M., Fjeldsa, J., Orlando, L., Barker, F. K., Jonsson, K. A., Johnson, W., Koepfli, K. P., O'Brien, S., Haussler, D., Ryder, O. A., Rahbek, C., Willerslev, E., Graves, G. R., Glenn, T. C., McCormack, J., Burt, D., Ellegren, H., Alström, P., Edwards, S. V., Stamatakis, A., Mindell, D. P., Cracraft, J., Braun, E. L., Warnow, T., Jun, W., Gilbert, M. T. P. & Zhang, G. (2014). Whole-genome analyses resolve early branches in the tree of life of modern birds. Science, 346, 1320–1331. https://doi.org/10.1126/science.1253451 Jombart, T., & Ahmed, I. (2011). adegenet 1.3-1: New tools for the analysis of genome-wide SNP data. Bioinformatics, 27, 3070–3071. https://doi.org/10.1093/bioinformatics/btr521 Kagawa, K. & Takimoto, G. (2018), Hybridization can promote adaptive radiation by means of transgressive segregation. Ecology Letters, 21: 264–274. doi:10.1111/ele.12891 Keller, I., Wagner, C. E., Greuter, L., Mwaiko, S., Selz, O. M., Sivasundar, A., Wittwer, S., Seehausen, O. (2013). Population genomic signatures of divergent adaptation, gene flow and hybrid speciation in the rapid radiation of Lake Victoria cichlid fishes. Molecular Ecology, 22, 2848– 2863. https://doi.org/10.1111/mec.12083 Kubatko, L. S., Degnan, J. H. & Collins, T. (2007). Inconsistency of phylogenetic estimates from concatenated data under coalescence. Systematic Biology, 56, 17–24. https://doi.org/10.1080/10635150601146041 Lamichhaney, S., Berglund, J., Almén, M. S., Maqbool, K., Grabherr, M., Martinez-Barrio, A., Promerová, M., Rubin, C., Wang, C., Zamani, N., Grant, B. R., Grant, P. R., Webster, M. T. & Andersson, L. (2015). Evolution of Darwin’s finches and their beaks revealed by genome sequencing. Nature, 518, 371–375. https://doi.org/10.1038/nature14181 Lamichhaney, S., Han, F., Berglund, J., Wang, C., Almén, M. S., Webster, M. T., Grant, R. G., Grant, P. R., & Andersson, L. (2016). A beak size locus in Darwin's finches facilitated character displacemnt during a drought. Science, 352, 470-474. https://doi.org/10.1126/science.aad8786 Lamichhaney S., Han, F., Webster, M. T., Andersson, L., Grant, B. R., & Grant, P. R. (2017). Rapid hybrid speciation in Darwin’s finches. Science. Doi: 10.1126/science.aao4593 Lanfear, R., Frandsen, P. B., Wright, A. M., Senfeld, T. & Calcott, B. (2016). PartitionFinder 2: new methods for selecting partitioned models of evolution for molecular and morphological phylogenetic analyses. Molecular Biology and Evolution 34:772-773. Lanfear, R., Calcott, B., Ho, S. Y. W. & Guindon, S. (2012). PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Molecular Biology and Evolution 29: 1695-1701. Leaché, A. D. (2009). Species tree discordance traces to phylogeographic clade boundaries in north American Fence lizards (Sceloporus). Systematic Biology, 58, 547–559. https://doi.org/10.1093/sysbio/syp057 Leaché, A. D., Banbury, B. L., Felsenstein, J., de Oca, A. & Stamatakis, A. (2015). Short Tree, Long


Tree, Right Tree, Wrong Tree: New Acquisition Bias Corrections for Inferring SNP Phylogenies. Systematic Biology, 64, 1032–1047. https://doi.org/10.1093/sysbio/syv053

Accepted Article

Leaché, A. D. & Oaks, J. R. (2017). The utility of single nucleotide polymorphism (SNP) data in phylogenetics. Annual Review of Ecology, Evolution, and Systematics 48. DOI: 10.1146/annurevecolsys-110316-022645. Lewis, P. O. (2001). A likelihood approach to estimating phylogeny from discrete morphological character data. Systematic Biology, 50, 913–925. https://doi.org/10.1080/106351501753462876 Lewontin, R. C., & Birch, L. C. (1966). Hybridization as a Source of Variation for Adaptation to New Environments. Evolution, 20, 315. https://doi.org/10.2307/2406633 Linkem, C. W., Minin, V. N., & Leaché, A. D. (2016). Detecting the anomaly zone in species trees and evidence for a misleading signal in higher-level skink phylogeny (Squamata: Scincidae). Systematic Biology, 65, 465–477. https://doi.org/10.1093/sysbio/syw001 Liu, K. J., Dai, J., Truong, K., Song, Y., Kohn, M. H., & Nakhleh, L. (2014). An HMM-Based Comparative Genomic Framework for Detecting Introgression in Eukaryotes. PLoS Computational Biology, 10. https://doi.org/10.1371/journal.pcbi.1003649 Maddison, W. P. (1997). Gene Trees in Species Trees. Systematic Biology, 46, 523. https://doi.org/10.2307/2413694 Mallet, J. (2005). Hybridization as an invasion of the genome. Trends in Ecology and Evolution, 20, 229–237. https://doi.org/10.1016/j.tree.2005.02.010 Mallet, J., Besansky, N., & Hahn, M. W. (2016). How reticulated are species? BioEssays, 38, 140–149. https://doi.org/10.1002/bies.201500149 Martin, C. H., Cutler, J. S., Friel, J. P., Dening Touokong, C., Coop, G., & Wainwright, P. C. (2015). Complex histories of repeated gene flow in Cameroon crater lake cichlids cast doubt on one of the clearest examples of sympatric speciation. Evolution, 69, 1406–1422. https://doi.org/10.1111/evo.12674 Martin, S. H., Davey, J. W., & Jiggins, C. D. (2015). Evaluating the use of ABBA-BABA statistics to locate introgressed loci. Molecular Biology and Evolution, 32, 244–257. https://doi.org/10.1093/molbev/msu269 Mastretta-Yanes, A., Arrigo, N., Alvarez, N., Jorgensen, T. H., Piñero, D., & Emerson, B. C. (2015). Restriction site-associated DNA sequencing, genotyping error estimation and de novo assembly optimization for population genetic inference. Molecular Ecology Resources, 15, 28–41. https://doi.org/10.1111/1755-0998.12291 McGuire, J. A., Linkem, C. W., Koo, M. S., Hutchison, D. W., Lappin, A. K., Orange, D. I., LemosEspinal, J., Riddle, B. R., & Jaeger, J. R. (2007). Mitochondrial introgression and incomplete lineage sorting through space and time: Phylogenetics of crotaphytid lizards. Evolution, 61, 2879–2897. https://doi.org/10.1111/j.1558-5646.2007.00239.x McVay, J. D., Hipp, A. L. & Manos, P. S. (2017). A genetic legacy of introgression confounds


phylogeny and biogeography in oaks. Proc. R. Soc. B. 284: 20170300. http://dx.doi.org/10.1098/rspb.2017.0300

Accepted Article

Meier, J. I., Marques, D. A., Mwaiko, S., Wagner, C. E., Excoffier, L., & Seehausen, O. (2017). Ancient hybridization fuels rapid cichlid fish adaptive radiations. Nature Communications, 8, 14363. https://doi.org/10.1038/ncomms14363 Melo-Ferreira, J., Seixas, F. A., Cheng, E., Mills, L. S., & Alves, P. C. (2014). The hidden history of the snowshoe hare, Lepus americanus: Extensive mitochondrial DNA introgression inferred from multilocus genetic variation. Molecular Ecology, 23, 4617–4630. https://doi.org/10.1111/mec.12886 Meyer, B. S., Matschiner, M., & Salzburger, W. (2016). Disentangling Incomplete Lineage Sorting and Introgression to Refine Species-Tree Estimates for Lake Tanganyika Cichlid Fishes. Systematic Biology, 66, 531-550. https://doi.org/10.1093/sysbio/syw069 Miao, B., Wang, Z., & Li, Y. (2016). Genomic Analysis Reveals Hypoxia Adaptation in the Tibetan Mastiff by Introgression of the Grey Wolf from the Tibetan Plateau. Molecular Biology And Evolution, 34, 734-743. https://doi.org/10.1093/molbev/msw274 Miller, M.A., Pfeiffer, W., and Schwartz, T. Creating the CIPRES Science Gateway for inference of large phylogenetic trees. In Proceedings of the Gateway Computing Environments Workshop (GCE), 14 Nov. 2010, New Orleans, LA. 1 - 8. Mishmar, D., Ruiz-Pesini, E., Golik, P., Macaulay, V., Clark, A. G., Hosseini, S., Brandon, M., Easley, K., Chen, E., Brown, M. D. Sukernik, R. I., Olckers, A. & Wallace, D. C. (2003). Natural selection shaped regional mtDNA variation in humans. Proceedings of the National Academy of Sciences, 100, 171–176. https://doi.org/10.1073/pnas.0136972100 Morales, A., Jackson, N., Dewey T., O'Meara, B. & Carstens, B. (2017) Speciation with gene flow in North American Myotis bats. Systematic Biology. 66, 440-452. https://doi.org/10.1093/sysbio/syw100 Novembre, J. & Stephens, M. (2008). Interpreting principal component analyses of spatial population genetic variation. Nature Genetics 40, 646-649. Doi: 10.1038/ng.139 Novikova, P. Y., Hohmann, N., Nizhynska, V., Tsuchimatsu, T., Ali, J., Muir, G., Guggisberg, A., Paape, T., Schmid, K., Fedorenko, O. M., Holm, S., Säll, T., Schlötterer, C., Marhold, K., Widmer, A., Sese, J., Shimizu, K. K.,Weigel, D., Krämer, U., Koch, M. A. & Nordborg, M. (2016). Sequencing of the genus Arabidopsis identifies a complex history of nonbifurcating speciation and abundant transspecific polymorphism. Nature Genetics, 48, 1077–1082. https://doi.org/10.1038/ng.3617 Paradis, E. (2010). Pegas: An R package for population genetics with an integrated-modular approach. Bioinformatics, 26, 419–420. https://doi.org/10.1093/bioinformatics/btp696 Parchman, T. L., Gompert, Z., Mudge, J., Schilkey, F. D., Benkman, C. W., & Buerkle, C. A. (2012). Genome-wide association genetics of an adaptive trait in lodgepole pine. Molecular Ecology, 21, 2991–3005. https://doi.org/10.1111/j.1365-294X.2012.05513.x Parsons, K. J., Son, Y. H., & Albertson, R. C. (2011). Hybridization Promotes Evolvability in African Cichlids: Connections Between Transgressive Segregation and Phenotypic Integration.


Evolutionary Biology, 38, 306–315. https://doi.org/10.1007/s11692-011-9126-7

Accepted Article

Patterson, N., Moorjani, P., Luo, Y., Mallick, S., Rohland, N., Zhan, Y., Genschoreck, T., Webster, T., & Reich, D. (2012). Ancient admixture in human history. Genetics, 192, 1065–1093. https://doi.org/10.1534/genetics.112.145037 Payseur, B. A., & Rieseberg, L. H. (2016). A genomic perspective on hybridization and speciation. Molecular Ecology, 25, 2337–2360. https://doi.org/10.1111/mec.13557 Pease, J. B., Haak, D. C., Hahn, M. W., & Moyle, L. C. (2016). Phylogenomics Reveals Three Sources of Adaptive Variation during a Rapid Radiation. PLOS Biology, 14, e1002379. https://doi.org/10.1371/journal.pbio.1002379 Pease, J. B., & Hahn, M. W. (2015). Detection and Polarization of Introgression in a Five-Taxon Phylogeny. Systematic Biology, 64, 651–662. https://doi.org/10.1093/sysbio/syv023 Pfeiffer, W., & Stamatakis, A. (2010). Hybrid mpi/pthreads parallelization of the raxml phylogenetics code. In Parallel & Distributed Processing, Workshops and Phd Forum (IPDPSW), 2010 IEEE International Symposium On, 208. Pickrell, J. K., & Pritchard, J. K. (2012). Inference of Population Splits and Mixtures from GenomeWide Allele Frequency Data. PLoS Genetics, 8, e1002967. https://doi.org/10.1371/journal.pgen.1002967 R Core Team. (2016). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from https://www.r-project.org/ Rambaut, A., Suchard, M., Xie, D., & Drummond, A. J. (2014). Tracer v1.6. Retrieved from http://beast.bio.ed.ac.uk/Tracer Reaz, R., Bayzid, M. S., & Rahman, M. S. (2014). Accurate phylogenetic tree reconstruction from quartets: A heuristic approach. PLoS ONE, 9, e0104008. https://doi.org/10.1371/journal.pone.0104008 Ree, R. H. & Hipp, A. L. (2015). Inferring phylogenetic history from restriction site associated DNA (RADseq). In ‘Next-Generation Sequencing in Plant Systematics’. Eds. E. Hörandl, M. S. Appelhans) Regnum Vegetabile, pp. 181–204. (International Association for Plant Taxonomy, IAPT: Königstein, Germany. Reich, D., Green, R. E., Kircher, M., Krause, J., Patterson, N., Durand, E. Y., Viola, B., Briggs, A. W., Stenzel, U., Johnson, P. L. F., Maricic, T., Good, J. M., Marques-Bonet, T., Alkan, C., Fu, Q., Mallick, S., Li, H., Meyer, M., Eichler, E. E., Stoneking, M., Richards, M., Talamo, S., Shunkov, M. V., Derevianko, A. P., Hublin, J., Kelso, J., Slatkin, M. & Pääbo, S. (2010). Genetic history of an archaic hominin group from Denisova Cave in Siberia. Nature, 468, 1053–1060. https://doi.org/10.1038/nature09710 Reich, D., Thangaraj, K., Patterson, N., Price, A. L., & Singh, L. (2009). Reconstructing Indian population history. Nature, 461, 489–94. https://doi.org/10.1038/nature08365 Rheindt, F. E., & Edwards, S. V. (2011). Genetic Introgression: An Integral but neglected component of speciation in birds. The Auk, 128, 620–632. https://doi.org/10.1525/auk.2011.128.4.620


Accepted Article

Rheindt, F. E., Fujita, M. K., Wilton, P. R., & Edwards, S. V. (2014). Introgression and phenotypic assimilation in zimmerius flycatchers (Tyrannidae): Population genetic and phylogenetic inferences from genome-wide SNPs. Systematic Biology, 63, 134–152. https://doi.org/10.1093/sysbio/syt070 Richards, E. J. & Martin, C. H. (2017). Adaptive introgression from distant Caribbean islands contributed to the diversification of a microendemic adaptive radiation of trophic specialist pupfishes. PLoS Genet 13: e1006919. https://doi.org/10.1371/journal.pgen.1006919 Ridgely, R. S., & Tudor, G. (1989). The Birds of South America. Vol. I. Austin, TX.: University of Texas Press. Scott, G. R., Schulte, P. M., Egginton, S., Scott, A. L. M., Richards, J. G., & Milsom, W. K. (2011). Molecular evolution of cytochrome c oxidase underlies high-altitude adaptation in the barheaded goose. Molecular Biology and Evolution, 28, 351–363. https://doi.org/10.1093/molbev/msq205 Schumer, M., Powell, D. L., Delclós, P. J., Squire, M., Cui, R., Andolfatto, P. & Rosenthal, G. G. (2017a). Assortative mating and persistent reproductive isolation in hybrids. Proceedings of the National Academy of Sciences. 201711238; DOI:10.1073/pnas.1711238114 Schumer, M., Xu, C., Powell, D., Durvasula, A., Skov, L., Holland, C., Sankararama, S., Andolfatto, P., Rosenthal, G. & Przeworski, M. Natural selection interacts with the local recombination rate to shape the evolution of hybrid genomes. Biorxiv, Posted November 1, 2017b. https://doi.org/10.1101/212407 Seehausen, O. (2004). Hybridization and adaptive radiation. Trends in Ecology and Evolution, 19, 198–207. https://doi.org/10.1016/j.tree.2004.01.003 Shen, X., Hittinger, C. T., & Rokas, A. (2017). Contentious relationships in phylogenomic studies can be driven by a handful of genes. Nature Ecology & Evolution, 1, 126. https://doi.org/10.1038/s41559-017-0126 Smith, S. a, Moore, M. J., Brown, J. W., & Yang, Y. (2015). Analysis of phylogenomic datasets reveals conflict, concordance, and gene duplications with examples from animals and plants. BMC Evolutionary Biology, 15, 150. https://doi.org/10.1186/s12862-015-0423-0 Solís-Lemus, C., & Ané, C. (2016). Inferring Phylogenetic Networks with Maximum Pseudolikelihood under Incomplete Lineage Sorting. PLoS Genetics, 12, e1005896. https://doi.org/10.1371/journal.pgen.1005896 Solís-Lemus, C., Yang, M., & Ané, C. (2016). Inconsistency of Species Tree Methods under Gene Flow. Systematic Biology, 65, 843–851. https://doi.org/10.1093/sysbio/syw030 Springer, M. S., & Gatesy, J. (2016). The gene tree delusion. Molecular Phylogenetics and Evolution, 94, 1–33. https://doi.org/10.1016/j.ympev.2015.07.018 Stamatakis, A. (2014). RAxML Version 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics, 1–2.


Accepted Article

Stankowski, S., & Streisfeld, M. A. (2015). Introgressive hybridization facilitates adaptive divergence in a recent radiation of monkeyflowers. Proceedings of the Royal Society of London B, 282, 20151666. https://doi.org/10.1098/rspb.2015.1666 Streicher, J. W., Devitt, T. J., Goldberg, C. S., Malone, J. H., Blackmon, H., & Fujita, M. K. (2014). Diversification and asymmetrical gene flow across time and space: Lineage sorting and hybridization in polytypic barking frogs. Molecular Ecology, 23, 3273–3291. https://doi.org/10.1111/mec.12814 Suh, A. (2016). The phylogenomic forest of bird trees contains a hard polytomy at the root of Neoaves. Zoologica Scripta, 45, 50–62. https://doi.org/10.1111/zsc.12213 Suh, A., Smeds, L., & Ellegren, H. (2015). The dynamics of incomplete lineage sorting across the ancient adaptive radiation of neoavian birds. PLoS Biology, 13, e1002224. https://doi.org/10.1371/journal.pbio.1002224 Swofford, D. (2002). PAUP*. Phylogenetic Analysis Using Parsimony (* and Other Methods). Version 4.0a147. Sunderland, Massachusetts: Sinauer Associates.

Than, C., Ruths, D & Nakhleh, L. (2008). PhyloNet: a software package for analyzing and reconstructing reticulate evolutionary relationships. BMC Bioinformatics 9: 322. http://doi.org/10.1186/1471-2105-9-322 Thom, G. Amaral, F. R., Hickerson, M. J., Aleixo, A., Araujo-Silva, L. E., Ribas, C. C., Choueri, E. & Miyaki, C. Y. (2018). Phenotypic and genetic structure support gene flow generating gene tree discordances in an Amazonian floodplain endemic species. Systematic Biology,, syy004, https://doi.org/10.1093/sysbio/syy004 Vargas, O. M., Ortiz, E. M., & Simpson, B. B. (2017). Conflicting phylogenomic signals reveal a pattern of reticulate evolution in a recent high‐Andean diversification (Asteraceae: Astereae: Diplostephium). New Phytologist. 214, 1736-1750. https://doi.org/10.1111/NPH.14530 VonHoldt, B. M., Cahill, J. A., Fan, Z., Gronau, I., Robinson, J., Pollinger, J. P., Shapiro, B., Wall, J. & Wayne, R. K. (2016). Whole-genome sequence analysis shows that two endemic species of North American wolf are admixtures of the coyote and gray wolf. Science Advances, 2, e1501714. https://doi.org/10.1126/sciadv.1501714 Wallbank, R. W. R., Baxter, S. W., Pardo-Diaz, C., Hanly, J. J., Martin, S. H., Mallet, J., Dasmahapatra, K. K., Salazar, C., Joron, M., Nadeau, N., McMillan, W. O. & Jiggins, C. D. (2016). Evolutionary Novelty in a Butterfly Wing Pattern through Enhancer Shuffling. PLoS Biology, 14, e1002353. https://doi.org/10.1371/journal.pbio.1002353 Warren, B. H., Bermingham, E., Bourgeois, Y., Estep, L. K., Prys-Jones, R. P., Strasberg, D., & Thébaud, C. (2012). Hybridization and barriers to gene flow in an island bird radiation. Evolution, 66, 1490– 1505. https://doi.org/10.1111/j.1558-5646.2011.01550.x Wen, D., Yu, Y., Hahn, M. W., & Nakhleh, L. (2016). Reticulate evolutionary history and extensive introgression in mosquito species revealed by phylogenetic network analysis. Molecular Ecology. 25, 2361-2372. https://doi.org/10.1111/mec.13544


Accepted Article

Wen, D., Yu, Y., & Nakhleh, L. (2016). Bayesian inference of reticulate phylogenies under the multispecies network coalescent. PLoS Genetics, 12, e1006006. https://doi.org/10.1371/journal.pgen.1006006 Winger, B. M. (2017). Consequences of divergence and introgression for speciation in Andean cloud forest birds. Evolution, 71, 1815-1831. https://doi.org/10.1111/evo.13251 Yu, Y., Degnan, J. H., & Nakhleh, L. (2012). The probability of a gene tree topology within a phylogenetic network with applications to hybridization detection. PLoS Genetics, 8, e1002660. https://doi.org/10.1371/journal.pgen.1002660 Yu, Y., Dong, J., Liu, K. J., & Nakhleh, L. (2014). Maximum likelihood inference of reticulate evolutionary histories. Proceedings of the National Academy of Sciences of the United States of America, 111, 16448–16453. https://doi.org/10.1073/pnas.1407950111 Yu, Y., Than, C., Degnan, J. H., & Nakhleh, L. (2011). Coalescent histories on phylogenetic networks and detection of hybridization despite incomplete lineage sorting. Systematic Biology, 60, 138– 149. https://doi.org/10.1093/sysbio/syq084

DATA ACCESSIBILITY: Mitochondrial DNA sequences are available on GenBank as accessions KT221117-KT221365,

and MH360743-MH360977. The raw GBS sequence data are accessible at the Nucleotide Short Read Archive, SRA accession: SRP146082. Sample localities are available in Table 1 and the Appendix; hyperlinks to Arctos database

record are provided when possible.

AUTHOR CONTRIBUTIONS: Designed the study: EJB, CCW. Collected samples: EJB, PMB, CCW. Sequenced samples: EJB,

PMB. Analyzed the data: EJB. Advised on analyses: PMB, ZAC CCW. Drafted manuscript: EJB, CCW. Improved manuscript: PMB, ZAC.


TABLE AND FIGURE CAPTIONS:

Accepted Article

TABLE 1. Genotyping-By-Sequencing Sampling. Elevation (Elev.) in meters. Reads with >30 mean Phred score are reported as Raw Reads; loci are from assembly B3. SetC and SetE columns indicate whether a sample was included in respective subsets. Abbreviations for museums are MSB: Museum of Southwestern Biology; FMNH: Field Museum of Natural History; USNM: Smithsonian Institute; UWBM: Burke Museum.

TABLE 2. ABBA/BABA and Four Population Test Results. Introgression test results based on Trees 131; each tree is arranged so Taxon1 and Taxon2 are sister, potentially introgressing taxa are Taxon2 and Taxon3, and Taxon 4 is closest to the root. Z-score and number of GBS unlinked, biallelic sites per test presented. Significant Z-scores are in bold; † indicates a Z-score that is >2.0, but did not pass false discovery threshold. Trees 1-31 were derived from RAxML results. Tree Topology indicates whether a tree is valid for the alternative phylogenies: SNAPP-S1 (S1), SNAPP-S2 (S2) and SVDquartet (SVD) trees. Set A includes all individuals; Set C includes a subset (Table 1). S. magellanicus abbreviations refer to central Peru (CP), southern Peru (SP), Peru (P), Bolivia (B), Argentina (A) or Bolivia and Argentina (BA).

Fig. 1. Principal component analysis of Spinus nDNA. Analyses were produced with 45,246 unlinked, variant sites from assembly B3. Individual birds are colored by taxon or clade as follows: S. crassirostris: maroon, S. atratus: black, S. magellanicus Peru: yellow, S. magellanicus Cochabamba, Bolivia: pink, S. magellanicus Argentina: orange, S. uropygialis: green. A. Includes all individuals. B. Excludes S. crassirostris and S. uropygialis. S. magellanicus Peru are colored as Southern Peru (light green) or Central Peru (yellow).


Fig. 2. Admixture results for Spinus nDNA. Analysis conducted with 45,246 unlinked, variant sites

Accepted Article

assembly B3; number of populations (k) varied from 2 to 6. A. CV error for k = 2 through 6; a low CV error indicates data are well described by corresponding k. B. Individual birds are color-coded by their a priori identification and, in the case of S. magellanicus, population of origin. Reconstructed populations that did not roughly correspond to a priori IDs were given gray-scale coloring.

Fig. 3. Maximum likelihood phylogeny of Spinus nDNA. Tree generated in RAxML with 45,246 concatenated loci from assembly B3. Branches colored by taxonomic units: S. cucullatus: red, S. crassirostris: maroon, S. atratus: black, S. magellanicus Peru: yellow, S. magellanicus Bolivia: pink, S. magellanicus Argentina: orange, S. uropygialis: green. Node support represents 500 bootstrap replicates.

Fig. 4. Summary of unrooted Andean Spinus tree topologies. The dashed red lines indicate alternate placements of outgroup S. cucullatus based on phylogenetic analyses. Branches are colored by taxon: S. crassirostris: maroon, S. atratus: black, S. magellanicus Peru: yellow, S. magellanicus Bolivia/Argentina: alternating orange and pink, S. uropygialis: green. A. Topology of the RAxML-B3 analysis; alternative rooting points are indicated for: the full concatenated RAxML analysis, labeled as RAxML “Best”; the SNAPP-S1 tree; and the RAxML SNP-only analysis without the Felsenstein invariant site correction. B. Topology of the SNAPP-S2 tree and the lineage and species trees from SVDquartets.

Fig. 5. mtDNA variation within Spinus. A. Maximum likelihood phylogeny based on ND3 (416 bp), ND2 (777 bp) and cytb (753 bp); colors correspond to S. crassirostris (maroon, 16 samples), S. uropygialis (green, 20), S. atratus (black, 8), S. cucullatus (red, 4) and S. magellanicus from Peru


(yellow, 43), Cochabamba, Bolivia (pink, 1), eastern Bolivia/northern Argentina (orange, 3). B. Map

Accepted Article

of S. magellanicus mtDNA for 100 specimens sampled in central and western South America; clades shown in A are indicated by color: crassirostris clade (maroon), atratus clade (black) or Argentina clade (orange).

Fig. 6. Summary of introgression results for alternative Spinus topologies. A. RAxML-B3 cladogram. B. SNAPP-S1 tree. C. SNAPP-S2 cladogram (S2). D. SVDquartet lineage tree; nodes with bootstrap support